Molten Salt-Shielded Synthesis (MS³) of MXenes in Air

1 Introduction

MXenes, a family of two-dimensional (2D) transition metal carbides and carbonitrides prepared from MAX phase precursors by selective etching of the A-site element, are promising materials for energy storage applications.^{[1, 2]} The general formula of MXenes is M_n+1X_nT_x (n = 1, 2, 3, or 4), where M is a transition metal(s), X is carbon and/or nitrogen, and T_x represents the surface terminations (–OH, –O, –F, –Cl, etc.).^[3] MXene synthesis is based on the selective etching of the A-layer of the corresponding MAX phase in F-containing solutions, such as HF or a LiF + HCl mixture.^[4] However, these methods pose significant risks to human safety and the environment owing to the formation of toxic and corrosive HF; thus, HF-free methods for MXene preparation are urgently needed. HF-free methods, such as electrochemical etching,^[5] molten salt etching,^[6] and hydrothermal routes,^[7] have been reported, but they require hours to days to prepare MXenes. In addition, most of the synthesis methods are only effective for etching Al-containing MAX phases, making them unsuitable for MAX phase precursors with other A-site elements (such as Si, Ga, and Zn). In 2020, a Lewis molten salt etching method was proposed for etching MAX phases with various A-site elements, which expands the potential range of MAX phase precursors.^[8] However, because of their oxidation sensitivity, MXenes are chemically unstable,^[9] particularly at high temperatures. Consequently, inert gas protection is required during synthesis that may cause complexity in synthesis procedures and experimental facilities, resulting in increased production costs.

In 2019, a molten salt-shielded synthesis (MS³) method was proposed to prepare various oxidation-sensitive materials, such as porous Ti, Ti₃SiC₂, and Ti₂AlN MAX phases, under an air atmosphere.^[10] In this process, low-melting-point KBr is used as a reaction medium to protect products from oxidation during high-temperature synthesis by forming a protective molten salt layer around the sample at low temperatures. Inspiring by the MS³ method of MAX phases synthesis and Lewis salt etching synthesis of MXenes, we recently proposed the preparation of MXenes by a one-pot molten salt synthesis method using M, A, and X (carbon) elements as precursors.^[11] Ti₃C₂T_x and Ti₂CT_x MXenes were obtained by two consecutive steps that occurred in the reactor: in-situ MAX phase synthesis at 1300 °C followed by in situ Lewis salt etching at 700 °C. This one-pot method enables the synthesis of carbide MXenes directly from elemental precursors for the first time, and the entire process was achieved in <8 h without inert gas protection. The one-pot synthesis strategy relies on the in-situ synthesis of MAX phases in molten salts, however, some MAX phases, such as Nb-based, most 413 and 514 MAX phases, have not been achieved by the molten salt method due to the severe volatilization of molten salts at the high synthesis temperatures near or beyond their boiling points,^{[12, 14]} which hinders the synthesis of corresponding MXenes (such as Nb₂C, Nb₄C₃, etc.) by one-pot synthesis method. Therefore, preparing MXenes by using MAX phases precursors is still indispensable by far.

In this study, the MS³ method was used to prepare MXenes with the corresponding MAX phases as precursors and the CuCl₂ Lewis salt as the etchant. Similarly, low-melting-point eutectic chlorides (NaCl + KCl) were employed as the salt bed to provide a molten salt environment and avoid direct contact between the reactants and the air atmosphere at high temperatures. Several carbide MXenes, as well as nitride MXenes that are typically difficult to be achieved by other methods, were successfully synthesized without inert gas protection in considerably less time. The scalability of the MS³ method was validated by etching large batches of 20 and 60 g of Ti₃AlC₂ MAX phase into the corresponding Ti₃C₂T_x MXene. When used as lithium storage anodes, the prepared MXenes exhibited superior lithium storage performance.

2 Results and Discussion

The X-ray diffraction (XRD) patterns of pristine Ti₃AlC₂ before and after molten salt etching for 10–40 min at 700 °C are shown in Figure 1b. After 20 min of etching, the diffraction peaks of the Ti₃AlC₂ MAX phase disappeared, while sharp diffraction peaks of the Ti₃C₂T_x MXene were observed.^[15] Even after 10 min of etching, the (00l) peaks of the Ti₃C₂T_x MXene remained noticeable. The (002) peak of the etched samples left shifted from 2θ = 9.49° to 7.94°, which corresponds to an increase in the interlayer spacing from 0.93 to 1.11 nm. Figure 1c–f present the scanning electron microscopy (SEM) images of the Ti₃C₂T_x MXenes etched for 10, 20, 30, and 40 min at 700 °C; all MXene particles exhibit the typical expected multilayered structure. Energy-dispersive X-ray spectroscopy (EDS) mapping (Figure S1) confirmed the homogenous distribution of Ti, C, Al, Cl, Cu, and O. The Al-to-Ti ratio decreased from 1:3 to 0.022:3 after 40 min of etching (Table S1), indicating the successful removal of Al from the pristine Ti₃AlC₂ MAX precursor.

Transmission electron microscopy (TEM) images show that the Ti₃C₂T_x MXene (Figure 1g) had a ribbon-like structure containing tens of restacked Ti₃C₂T_x atomic layers, which explains well the sharp XRD peaks. Atomically resolved high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images along the [11$\overline{2}$0] and [1$\overline{1}$00] directions provide the atomic structure information of the Ti₃C₂T_x layers. Figure 1h,i clearly show the restacking of Ti₃C₂T_x MXene layers, with each Ti₃C₂T_x MXene layer containing five atomic layers: the two outer atomic layers are Cl atoms, and the three core atomic layers are Ti atoms. This observation matches well with previous results reported for Cl-terminated Ti₃C₂T_x MXenes.^{[11, 16, 17]} The observed structure is completely different from the pristine Ti₃AlC₂ MAX phase precursor because no Al atomic layer was observed between the Ti₃C₂ layers, confirming the removal of Al by chemical etching.

Overviews of the X-ray photoelectron spectroscopy (XPS) spectra of the Ti₃AlC₂ precursor (black) and Ti₃C₂T_x MXene (red) are shown in Figure S2. While the Ti, C, and O signals are present for both the Ti₃AlC₂ MAX precursor and Ti₃C₂T_x MXene samples, the Cl signal is only detected for the Ti₃C₂T_x MXene, which confirms the presence of Cl surface groups. Moreover, the Ti–Al bond signal at 73.5 eV and Al signal at 71.5 eV are not visible for the Ti₃C₂T_x MXene (Figure S2b), confirming the removal of the Al layer.^{[18, 19]} The Al–O peak could be caused by a small hydrolysis product of AlCl₃, as reported previously,^[20] or Al₂O₃ impurities from the precursors.^[21] The Ti 2p XPS spectrum of the Ti₃C₂T_x MXene is shown in Figure S3a. The peaks at 454.5 and 456.0 eV are assigned to Ti–C (I) (2p_3/2) and Ti–C (II) (2p_3/2) bonds, while the peaks at 460.4 and 461.8 eV correspond to Ti–C (I) (2p_1/2) and Ti–C (II) (2p_1/2) bonds.^{[22, 23]} In addition, the peaks at 458.4, 464.3, 457.9, and 463.6 eV can be associated with Ti–O (2p_3/2), Ti–O (2p_1/2), Ti–Cl (2p_3/2), and Ti–Cl (2p_1/2) bonds,^{[20, 22, 24]} respectively. The Ti–O and Ti–Cl chemical bonds confirm the presence of –Cl and –O surface groups on the Ti₃C₂T_x MXene. Furthermore, the existence of –O surface groups is verified by the fit C 1s (Figure S3b) and O 1s (Figure S3c) spectra. Additional information on the XPS analysis is shown in the Supporting Information (Figure S3 and Table S2).

In the next step, the synthesis process was scaled by etching large batches (20 and 60 g) of the Ti₃AlC₂ MAX phase. The XRD patterns (Figure 2a) of the Ti₃C₂T_x prepared from 0.5, 20 and 60 g batches exhibited no differences, with the same characteristic peaks located at the same 2θ angles. The SEM images in Figure 2b–d show that the structural morphology of the Ti₃C₂T_x MXene prepared from the large batches was almost identical to that of the MXene from the small batch of 0.5 g, demonstrating the scalability of the MS³ method. This argument is also validated by XPS analysis. These characterizations confirm no distinct structural or chemical state differences between the samples produced in 0.5 g and 20 g batches (Figure S3). Hence, we believe that this MS³ method can be scaled for the industrial production of Ti₃C₂T_x MXenes owing to its simplicity. In addition, the yield rate of Ti₃C₂T_x MXene was calculated. Starting with 20 g of Ti₃AlC₂ MAX, 14.46 g of Ti₃C₂T_x MXene was obtained after washing with APS, giving a yield rate of 72.3%.

In addition to its simplicity and scalability, the MS³ route allows for the preparation of various MXenes, including Ti₂CT_x, Ti₃CNT_x, and Ti₄N₃T_x. Figure 3a–c show that all characteristic peaks of the MAX phases disappeared from the XRD patterns, while the (00l) peaks shifted to lower 2θ angles after etching, indicating the successful synthesis of the corresponding MXenes. Furthermore, SEM images (Figure 3d–f) show multilayered MXenes. Many small particles were observed on the MXene particle surfaces, particularly on the Ti₂CT_x MXene sample, which could be explained by aluminum oxide, as discussed in the XPS analysis. EDS maps (Figures S4–S6) show the homogeneous distribution of elements, including Ti, Al, C, N, Cl, Cu, and O, and particle-like aggregation of Al and O, confirming the presence of aluminum oxide particles on the surface of MXenes. The Al and Cu contents are considerably low in the MXenes (Table S1), indicating the removal of Al by chemical etching and the removal of Cu by APS washing. TEM images (Figure 3g–i) confirm the multilayered structure of the prepared MXenes, in good agreement with previously reported results.^{[8, 11]} In addition, Figures S7–S9 present the XPS spectra of Ti₄N₃T_x, Ti₃CNT_x, and Ti₂CT_x, respectively. All XPS spectra confirm the removal of Al in the MAX phases and the successful synthesis of the corresponding MXenes. More information is provided in Tables S3–S5. Note that nitride MXenes are typically difficult to obtain using most MXene synthesis methods. For example, the Ti₄N₃T_x MXene has only been synthesized using molten fluoride salts.^{[6, 25]} Thus, the MS³ method has proven to be a general method for preparing both carbide and nitride MXenes. The yield rates of the Ti₂CT_x, Ti₃CNT_x, and Ti₄N₃T_x MXenes from the corresponding MAX phases were 65%, 78%, and 70%, respectively.

The electrochemical properties of the MS³-MXenes were investigated in half (coin) cells with MXenes as working electrodes and lithium foils as counter and reference electrodes. Cyclic voltammetry (CV) profiles of various MXenes are shown in Figure S10 at scan rates from 0.5 to 20 mV s⁻¹. As shown, Ti₄N₃T_x is considerably less promising as a lithium storage anode because of its lower specific capacity and higher delithiation potential (Figure S10d). In the CV curves of the Ti₂CT_x, Ti₃CNT_x, and Ti₃C₂T_x electrodes (Figure 4a), no noticeable redox peaks were observed, as typically achieved with battery-type materials, and a constant current was obtained in a broad potential range, particularly for the delithiation process; this is similar to the intercalation pseudocapacitive behavior reported for Nb₂O₅ in Li⁺-containing nonaqueous electrolytes.^[26] In addition, the electrochemical signatures agree well with those of a recently reported molten salt prepared MXene (MS-MXene).^[8] Figure 4b presents the galvanostatic charge-discharge curves of the Ti₂CT_x, Ti₃CNT_x, and Ti₃C₂T_x MXenes at 0.1 A g^-1. A maximum reversible specific capacity of 280 mAh g⁻¹ was delivered with the Ti₂CT_x MXene electrode, which outperformed the maximum capacity of MS-Ti₂CT_x.^[8] The rate performances of Ti₂CT_x, Ti₃CNT_x, Ti₃C₂T_x, and Ti₄N₃T_x at different specific currents were tested (Figure S11) after pre-cycling at a current density of 0.1 A g⁻¹, and the results are summarized in Tables S6–S9. A maximum specific capacity of 295 mAh g⁻¹ was achieved with the Ti₂CT_x MXene electrode at 0.05 A g⁻¹. When the discharge rate was increased to 0.1, 0.2, 0.5, 1.0, 2.0, and 5 A g⁻¹, the Ti₂CT_x electrode retained capacities of 280, 260, 230, 200, 165, and 115 mAh g⁻¹, respectively. The Ti₃C₂T_x and Ti₃CNT_x electrodes had lower specific capacities than that of the Ti₂CT_x electrode, and further study is needed to understand the charge storage mechanism. Figure 4c compares the specific capacities of the MXene electrodes prepared in this work (MS³-MXene) with those of MXene electrodes obtained by HF etching and Lewis salt etching methods at different C rates. The MS³-MXenes achieved a higher capacity and improved power performance for lithium-ion storage compared with those of MS- and HF-prepared MXenes.^{[8, 27, 28]} Figure 4d shows the cycling stability of the MXene electrodes at a specific current of 1 A g⁻¹. The cycling capacities remained stable for 1000 cycles, indicating that these MXene electrodes have good cycling properties. These results demonstrate that MS³-MXene electrodes possess excellent electrochemical performance for Li-ion storage in nonaqueous electrolytes.

3 Conclusion

We report the MS³ method for preparing MXenes in the air using low-melting-point eutectic chlorides as a reaction salt bed and Lewis-acid salts as etchants. The molten chloride salt bed prevents direct contact between the MAX precursor and air during the high-temperature etching reaction. Several types of MXenes were successfully prepared using the MS³ method, and its large-scale production capacity was also highlighted. Moreover, the prepared MS³-MXenes achieved excellent electrochemical performance for lithium-ion storage. Compared to other MXene synthesis methods, the MS³ method does not use any hazardous reactants or inert gas protection and allows for the fast synthesis of both carbide and nitride MXenes. It can be easily scaled to etch 60 g of the MAX phases. We believe that the proposed MS³-MXene synthesis method is an important advance that will be of great interest to MXene scientific researchers and industry.

4 Experimental Section

Conflict of Interest

The authors declare no conflict of interest.